摘要
为改善支持向量机的性能,从深度学习的角度研究核学习的方法,提出了基于多层感知器的深度核映射支持向量机模型(deep kernel mapping support vector machine,DKMSVM)以及相应的学习算法.该模型首先通过多层感知器学习一个从原始输入空间到合适维度空间的核映射代替传统意义上的核函数,然后直接在合适维度空间使用支持向量机进行分类,而不是采用核技巧进行求解.实验结果验证了DKMSVM的有效性.
To improve the performance of support vector machines （ SVMs）,from the deep learning 爷 s point of view,a kernel learning method was studied and a deep kernel mapping support vector machine （DKMSVM ） was proposed based on multi-layer perceptron together with the corresponding learning algorithm. Firstly,a kernel mapping from the original input space to a proper dimensional space through a multilayer perceptron instead of a traditional kernel function was researched in this model. Then a SVM was used to classify in the proper dimensional space without kernel tricks. Experimental results demonstrate the effectiveness of DKMSVM.
出处
《北京工业大学学报》
CAS
CSCD
北大核心
2016年第11期1652-1661,共10页
Journal of Beijing University of Technology
基金
国家自然科学基金资助项目(61175004)
高等学校博士学科点专项科研基金资助项目(20121103110029)
关键词
核学习
深度学习
多层感知器
支持向量机
kernel learning
deep learning
multi-layer perceptron
support vector machine
